TSpace Research Repository tspace.library.utoronto.ca

Structural, optical, and electronic studies of wide-bandgap lead halide perovskites

Riccardo Comin, Grant Walters, Emmanuel Thibau,

Oleksandr Voznyy, Zheng-Hong Lu, Edward H. Sargent

Version Post-Print/Accepted Manuscript

Citation (published version)

Comin, R., Walters, G., Thibau, E. S., Voznyy, O., Lu, Z.-H., & Sargent, E. H. (2015). Structural, optical, and electronic studies of wide-bandgap lead halide perovskites. J. Mater. Chem. C, 3(34), 8839–8843. doi:10.1039/c5tc01718a

Publisher’s Statement The final published version of this paper is available at Royal Society of Chemistry via http://dx.doi.org/10.1039/c5tc01718a.

How to cite TSpace items

Always cite the published version, so the author(s) will receive recognition through services that track citation counts, e.g. Scopus. If you need to cite the page number of the TSpace version (original manuscript or accepted manuscript) because you cannot access the published version, then cite the TSpace version in addition to the published version using the permanent URI (handle) found on the record page.

http://dx.doi.org/10.1039/c5tc01718a

Structural, optical, and electronic studies of wide-bandgap lead

halide perovskites

Riccardo Comin†,a, Grant Walters†,a, Emmanuel Thibaub, Oleksandr Voznyya, Zheng-Hong Lub,

Edward H. Sargenta,*

a) Department of Electrical and Computer Engineering, University of Toronto. 10 King’s

College Road, Toronto, Ontario, M5S 3G4, Canada.

E-mail: ted.sargent@utoronto.ca

b) Department of Materials Science and Engineering, University of Toronto. 184 College

Street, Toronto, Ontario M5S 3E4, Canada.

† These authors contributed equally

ABSTRACT We investigated the family of mixed Br/Cl organolead halide perovskites, which

enable light emission in the blue-violet region of the visible spectrum. We report the structural,

optical and electronic properties of this air-stable family of perovskites, demonstrating full

bandgap tunability in the 400-550 nm range and enhanced exciton strength upon Cl substitution.

We complement this study by tracking the evolution of the band levels across the gap, thereby

providing a foundational framework for future optoelectronic applications of these materials.

Introduction

Organometallic lead halide thin films have been attracting an ever-growing interest from the

photovoltaic scientific community following breakthroughs in solar-to-electric power conversion

efficiencies,1,2 which have recently risen to 20.1%3 in a remarkably short timeframe. While the

predominant research thrusts have since been oriented towards the use of these materials as

active layers in light-harvesting devices, more recent studies have demonstrated optoelectronic

applications of perovskite thin films in light-emitting-devices (LEDs),4,5 lasers,6–8

photodetectors,9 and photocatalysts.10 The physical chemistry of these compounds is particularly

tolerant to mixed-halide synthesis2,11, which can lead to pure or mixed phases according to the

structural compatibility of the different anions12–17, an aspect which is key to the impressive

versatility characterizing this class of materials. While wide-bandgap perovskites are not suitable

for photovoltaic applications due to their inability to their transparency to most of the solar

spectrum, the LED technology can fully leverage the optical tunability and spectral purity that

can be simultaneously engineered in Br- and Cl-based lead halide perovskites. The optical

tunability, in particular, is a direct consequence of the negative electronic compressibility of

these systems, whereby a compression of the unit cell – implemented via anion substitution – can

be used to increase the bandgap over a wide spectral range. In this context, recent studies have

shown promising prospects for the use of lead halide perovskite films in light emission devices

whose spectral response can be chemically tuned throughout the entire visible spectrum.4,5 While

extensive efforts are underway to optimize the efficiency of red and green perovskite LEDs, Cl-

based films for blue-violet light emission have only been studied to a somewhat peripheral

extent,5,6 partly due to the lack of information on the electronic band levels. While earlier works

have preliminarily addressed certain aspects of the near-gap bandstructure and related optical

response in Cl- and Br- based thin films,18,19 a detailed and systematic study of the electronic

structure in the entire range of halide substitution is, to date, still lacking.

We therefore sought to shed light on the electronic structure of wide-bandgap MAPbBr3-xClx thin

film perovskites. Here we investigate the evolution of the structural and optical properties, as

well as of the valence and conduction band levels, as a function of Cl/Br substitution using x-ray,

optical, and photoelectron methods.

Results and discussion

For this study, we used solution processing routes to synthesize thin films of methylammonium

(MA = CH3NH3) lead halide perovskite MAPbBr3-xClx, with Cl concentration values 𝑥 =

0, 0.6, 1.2, 1.8, 2.4, 3 (see Supplementary Information for more details). For the pure-Cl

compound, different chemical precursors have been used during synthesis, resulting in films

showing no appreciable differences in the properties discussed hereafter (a comparison of the

surface morphology is presented Figure S1, Supplementary Information).

Structural properties

For the study of structural properties we used x-ray diffraction (XRD), with corresponding scans

shown in Figure 1a. Multiple reflections can be seen that are associated with the cubic 𝑃𝑚3̅𝑚

space group. The lack of certain Bragg peaks for 𝑥 ≥ 1.8 is likely due to a higher degree of

orientation in the film, confirmed by the presence of flat-lying cuboids in MAPbCl3 (see again

Figure S1, Supplementary Information). Most remarkably, we find a smooth evolution of the

lattice structure across the entire concentration range, as demonstrated by the monotonic

variation in the angular position of all Bragg reflections. The inset of Figure 1a showcases a

zoomed view of the (001) Bragg peak as a function of x, which implies a contraction of the

perovskite cage as the smaller anion Cl is substituted for Br. The cubic lattice constant 𝑎 is

extracted from the angular position of the (001) Bragg reflection 𝜃𝐵 using 𝜆 2𝑎⁄ = sin(𝜃𝐵 2⁄ ),

and the average (shortest) grain dimension 𝐷 is calculated from 𝐷 = 0.9 ∙ 𝜆 (Δ𝜃𝐵 ∙ cos 𝜃𝐵)⁄

where Δ𝜃𝐵 is the Bragg peak linewidth and 𝜆 is the x-ray wavelength (see Figure S2,

Supplementary Information). The lattice constant is plotted in Figure 1b together with the

intensity of the Br-3d core levels measured with x-ray photoemission spectroscopy (XPS) (see

also Figure S3, Supplementary Information). The monotonic trend in the lattice constant, which

increases by ~5% from MAPbBr3 to MAPbCl3, confirms the progressive expansion of the unit

cell. The variation over time of the sample structure and crystallinity is presented in Figure 1c

for the case of MAPbCl3. The negligible evolution of the peak intensity and broadening

combines with the absence of any sign of decomposition (inferred from the lack of peaks from

the metal precursor PbCl2) to provide evidence for an excellent structural stability in ambient

conditions over more than 5 months.

Optical response

We then proceeded to study the optical properties of MAPbBr3-xClx films using absorption and

photoluminescence (PL) spectroscopy. Figure 2a shows a sequence of thin films on glass with

compositions varying from pure Br- to pure Cl-perovskite, and where the change in the optical

response is clearly discernible. The change in color indicates the increase in the bandgap as Br is

replaced with Cl, consistent with the trend observed in the Br/I mixed halide perovskites,20 and

more generally expected due to the reduced covalency of the lead-halide bonds within the

network of Pb-X6 (X = Cl, Br, I) corner-sharing octahedra. Optical absorbance spectra, as well as

PL scans, have been measured for films deposited on Spectrosil® substrates (the latter enable an

extended spectral range of perovskite film investigation thanks to their high-energy absorption

onset at 𝜆~200 nm) and are plotted in Figure 2b. The gradual evolution in the absorption edge

is consistent with the continuous structural change observed with XRD, while at the same time

the excitonic pre-peak is found to increase in intensity, a behavior which is related to an

increasing exciton binding strength as discussed in more detail below. Consistent with the high

degree of stability seen using XRD, the optical response remains unaltered over extended periods

of time and up to 1 month after fabrication (see Figure S4, Supplementary Information),

confirming the overall stability of these materials in ambient conditions. Figure 2c shows the

normalized PL spectra, featuring a complementary progression in the emission wavelength,

which blue-shifts from the green-emitting MAPbBr3 to the deep-violet MAPbCl3, thus

confirming the controllable color tunability upon halogen substitution. The raw PL data, as well

as the PL quantum efficiency (defined as the ratio of emitted to absorbed photons), are shown in

Figure S5, Supplementary Information.

The absorbance profile, shown in Figure 2d for the case of MAPbCl3, is comprised of two main

contributions: an excitonic peak at lower energies, and an extended absorption edge representing

the continuum of valence-to-conduction band electronic excitations. Equation 1 is used to model

the excitonic absorption term using an asymmetric lineshape (to account for the higher-lying

exciton levels as well as the interaction between the latter and the particle-hole continuum21) in

the form of a hyperbolic secant:

Figure 1. a) X-ray diffraction scans of Pb(Ac)2-based thin films of MAPbBr3-xClx (continuous lines) on fluorine-doped tin oxide (FTO). Shaded area: background signal from a pure FTO substrate (corresponding peaks are marked with a star). Inset: magnified view of the (100) diffraction peak. b)

Evolution of the lattice parameter and Br-3d peak intensity vs. x. c) XRD pattern of PbCl2-based MAPbCl3 over time, demonstrating structural

stability of pure-Cl films in atmosphere. The bottom bars mark the angular position of the powder XRD peaks in PbCl2.

𝛼𝑒𝑥𝑐(𝐸) = 𝐼𝑒𝑥𝑐 ∙ 𝐸−1 ∙ [exp (

𝐸−𝐸𝑒𝑥𝑐

Γ(𝐸−𝐸𝑒𝑥𝑐)) + exp (

𝐸𝑒𝑥𝑐−𝐸

Γ(𝐸−𝐸𝑒𝑥𝑐))]

−1

(1)

where I𝑒𝑥𝑐 and E𝑒𝑥𝑐 represent the exciton intensity and energy respectively, and Γ(𝐸) = 2 ∙

Γ𝑒𝑥𝑐 ∙ (1 + 𝑒𝑎𝐸)−1 is a generalized, energy-dependent linewidth. The exciton energy is related to

the exciton binding energy 𝐺 by E𝑒𝑥𝑐 = E𝑔 − 𝐺 𝑛2⁄ (E𝑔 is the electronic bandgap) and here we

use a single line corresponding to the exciton ground state (𝑛 = 1). For the main absorption

edge, we use a modified version22 of the canonical lineshape for direct transitions23 (Equation

2):

𝛼𝑐𝑜𝑛𝑡(𝐸) = 𝐼𝑐𝑜𝑛𝑡 ∙ √𝐺 ∙ 𝐸−1 ∙ [1 − exp(−𝜋√𝐺 𝐸 − 𝐸𝐺⁄ )]

−1∙ ℎ(𝐸) ∙ 𝜃(𝐸 − 𝐸𝑔) (2)

where 𝐼𝑐𝑜𝑛𝑡 represents the intensity of the absorption edge, 𝜃(𝐸) is the Heaviside step function,

and the function ℎ(𝐸) = [1 − 𝑏(𝐸 − 𝐸𝐺)]−1 parametrizes the deviation from the parabolic band

regime (following Ref. 22). The expression for the absorption edge in Equation 1 is subsequently

convoluted with a hyperbolic secant function with constant linewidth Γ𝑐𝑜𝑛𝑡. The experimental

data are fitted with the sum of the excitonic and continuum terms, and the resulting profile

(together with the individual components) is shown in Figure 2d (for the other compositions see

Figure S6, Supplementary Information). In general, we make use of the fact that the same

parameter 𝐺 controls both the position of the exciton line and the lineshape of the main

absorption edge in order to provide a reliable estimate for the bandgap as well as the exciton

binding energy. The results for the bandgap, the Stokes shift (defined as E𝑒𝑥𝑐 − 𝐸𝑃𝐿, where 𝐸𝑃𝐿

is the PL peak energy), and the exciton binding energy 𝐺 are reported in Figure 2e, as well as in

Table 1. As already noted from the absorption spectra, 𝐺 increases with Cl content, which

reflects the more pronounced ionic character in Cl-based perovskites. In pure MAPbCl3 we

determine an exciton binding energy of about 50 meV, more than double the value for MAPbBr3

(𝐺~21 meV), while in general our

absolute values are compatible or slightly

lower than previous optical studies on I-

based perovskite.22,24

Electronic structure

For a mapping of conduction and valence

band levels across the bandgap, we have

further investigated the energetics of

occupied states using Ultraviolet

Photoemission Spectroscopy (UPS).

Photoemission not only measures the

binding energy (𝐸𝐵) of valence electrons

with respect to the chemical potential, but

also provides a means for determining the

work function (𝜙) of the material which

sets the minimum kinetic energy (𝐸𝐾) required for photoelectrons to escape the film according to

𝐸𝐾 = ℎ𝜈 − 𝜙 − |𝐸𝐵|, where ℎ𝜈 is the photon energy. Therefore, since 𝐸𝐾 must be positive, a

drop in the photoelectron signal in the form of an intensity cutoff occurs at |𝐸𝐵| = ℎ𝜈 − 𝜙,

which is clearly visible in the photoelectron traces shown in Figure 3a, and measured using a

UV source (ℎ𝜈 = 21.2 eV). Linear fits to the pre- and post-cutoff regions are used to estimate

the energy onset from which the work function can be derived (values listed in Table 1). The

valence band photoemission spectra are displayed in Figure 3b, with magnified view near the

Fermi level shown in the inset plot. Again, linear fits below and above the onset of the occupied

Figure 2 a) Photograph of Br/Cl mixed films on glass, visibly showing the

progressive change in optical properties. b,c) Optical absorbance and photoluminescence spectra (370 nm excitation) of MAPbBr3-xClx films on

Spectrosil®. d) Plot of the absorbance (blue markers) of MAPbCl3 as a function

of energy with two-component (exciton and band-to-band absorption) fit overlaid (black line). e) Evolution of the electronic bandgap (grey circles),

Stokes shift (blue diamonds), and exciton binding energy (orange squares) as

a function of Cl content x.

density of states are used to locate the valence band maximum, as measured with respect to the

Fermi energy 𝐸𝐹. The resulting band structure vs. Cl concentration is summarized in the

schematic presented in Figure 3c (all energies are referred to the vacuum level). A close

inspection of this diagram leads to the following considerations: (i) the pure-Br perovskite

exhibits an ostensible asymmetry in the position of the band edges, consistent with previous

findings;25 (ii) when Cl is substituted for Br, the material develops a more intrinsic electronic

structure, which might suggest a cleaner bandgap with lower defect density; (iii) for increasing

Cl content, the valence band position shows little variation, while a monotonic trend is found for

the conduction band edge (with the exception of 𝑥 = 0.6); (iv) the work function does not follow

a specific trend with halide substitution, and in all cases ranges around 5 eV.

Conclusions

Thin films of mixed-halide wide-bandgap perovskites reveal a number of interesting

characteristics. Structural measurements show that Cl can be seamlessly substituted for Br,

attributable to the lattice compatibility of the pure-halide compounds. The smooth structural

Figure 3 a) Low-energy secondary electron cutoff of MAPbBr3-xClx films as measured using x-ray photoemission; arrows are located at the intersection

between linear fits (orange traces) and mark the cutoff position in each spectrum. b) Valence band photoemission spectra near the Fermi energy 𝐸𝐹 with position of the valence band maximum pinned at the onset in the photoelectron signal as determined by linear fits (orange traces). In the main graph, red triangles and white circles mark the positions of the valence band onset and of the lowest-energy peak, respectively. c) Schematics of the electronic structure

of MAPbBr3-xClx films across the gap, with work function (𝜙), valence band maximum (VBM), and conduction band minimum (CBM) values reported as the energy distance from the vacuum level.

evolution translates to a full tunability of optical properties, a behavior which enables the

fabrication of spectrally-tunable lead halide perovskites having bandgaps in the 400 to 550 nm

wavelength range. In light of the promising radiative performance of I-based perovskites,4 the

increased ionicity of Cl-containing perovskites suggests that the latter might represent a

compelling platform for efficient, air-stable, light-emitting devices in the blue-violet region of

the visible spectrum. Our photoemission results ultimately provide a comprehensive mapping of

the electronic levels to guide future device design strategies in these materials.

Experimental

Synthesis of lead halide perovskite thin films

Metal salt precursors Pb(Ac)2 (Ac = CH3COO), PbBr2 and PbCl2 (99.999% purity) were

purchased from Sigma Aldrich. The organic salt precursor CH3NH3Br (CH3NH3Cl) was

synthesized by the reaction of methylamine aqueous solution and the corresponding

hydrobromide (hydrochloride) acid in a round-bottom flask at 0 °C for 2 h with stirring. The

precipitate was recovered by placing the solution on a rotary evaporator and carefully removing

the solvents. The raw product CH3NH3X was re-dissolved in absolute ethanol and precipitated

with the addition of diethyl ether for recrystallization. After filtration, the previous step was

repeated again. Finally, the solid product was collected and dried at 60 °C in a vacuum oven for

24 h.

All substrates were chemically washed prior to film deposition. Substrates were immersed and

sonicated in baths of Triton X-100 cleaning solution, isopropanol, and deionized water

sequentially. After each bath, substrates were rinsed with deionized water. The substrates were

then dried in an oven at 90 °C for 1.5 to 2 hours. This was followed by a final soak in an

isopropanol bath for six hours and subsequent drying with nitrogen gas. After chemical washing,

glass and Spectrosil® substrates were plasma treated for five minutes with an oxygen plasma in

order to improve film wetting. Following plasma treatment, films were immediately deposited on

the substrates using a one-step or a two-step method.

For one-step syntheses, lead precursor [Pb(Ac)2, PbBr2, or PbCl2] and methylammonium halide

were weighed and added together in a vial such that the molar ratio of lead to halide was 1:3.

Solvent was then added to the powder precursors. For two-step syntheses, lead precursor and

methylammonium halide were weighed out separately and solvent was added to each powder

individually. Unless stated otherwise, dimethyl sulfoxide (DMSO) was used as the solvent and

the concentration of Pb2+ was 0.6 M/L in all experiments. Solutions were vortexed and then

sonicated for ~20 min. For one-step syntheses, 150 𝝁L of solution were deposited on clean

substrates for spin-coating. Samples were spun at 4000 rpm for 1 minute. For two-step syntheses,

ten drops of lead precursor solution were spun using the same procedure; spin coating was then

repeated with ten drops of methylammonium halide solution. In the case of anti-solvent rapid

crystallization, a 50 𝝁L aliquot of chlorobenzene was deposited onto the spinning sample after

spinning at 10000 rpm for 10 seconds and 30 seconds at 5000 rpm; spinning was continued for

30 seconds at 5000 rpm after chlorobenzene addition. All samples were immediately annealed

for 45 minutes at 100 °C on a hotplate. The entire fabrication procedure was performed in a

room-temperature nitrogen atmosphere.

Characterization of perovskite thin films

X-ray diffraction (XRD). XRD scans have been acquired on a Rigaku MiniFlex 600

diffractometer equipped with a NaI scintillation counter and using monochromatized Cu-𝐾𝛼

radiation (𝜆 = 1.5406 Å), with a detector angle (2𝜃) step of 0.03°. The angular resolution is

∆2𝜃~0.1°, as determined from diffraction on LaB6 single crystals.

X-ray photoelectron spectroscopy (XPS). The Cl concentration values in mixed Br/Cl

perovskite films have been obtained from the analysis of XPS spectra. The latter have been

measured using a PHI-5500 setup using a monochromatized Al-𝐾𝛼 x-ray anode (1486.7 eV) to

excite photoelectrons in an ultrahigh vacuum atmosphere (~10−9 Torr). The binding energy

scale was calibrated using the Au-4𝑓7/2 peak (83.98 eV) and the Cu-2𝑝3/2 peak (932.67 eV) of

sputter-cleaned Cu.

Scanning electron microscopy (SEM). A Quanta FEG 250 environmental scanning electron

microscope was used to collect SEM images. Images were acquired with an Everhart-Thornley

secondary electron detector, under high vacuum, and with a 10.0 kV accelerating voltage. Cross-

sectional samples were lightly carbon-coated prior to imaging in order to prevent sample

charging.

Optical absorption: Optical absorption scans have been acquired with a PerkinElmer Lambda

950 UV/Vis/NIR spectrophotometer equipped with a 150 mm integrating sphere, in the

wavelength range 250 − 900 𝑛𝑚 and using 2 𝑛𝑚 incremental steps.

Photoluminescence (PL) spectra. PL emission profiles have been measured using a Horiba

Fluorolog setup equipped with a monochromatized Xe lamp and with a photon detection unit

featuring a double-grating spectrometer and a photomultiplier tube covering the spectral range

300 − 850 𝑛𝑚. All PL datasets have been acquired using 𝜆 = 370 nm excitation and in

reflection geometry (at an incidence angle of 30° from the sample surface, to avoid spurious

reflections of the incident light into the spectrometer). The raw data have been subsequently

corrected for the experimentally-determined spectral sensitivity of the photon detector.

Ultraviolet photoemission spectroscopy (UPS). Photoelectron Spectroscopy (PES) was

performed in a PHI5500 Multi-Technique system using monochromatic Al-𝐾𝛼 radiation (XPS,

ℎ𝜈 = 1486.7 eV) and non-monochromatized He-𝐼𝛼 radiation (UPS, ℎ𝜈 = 21.22 eV). All

workfunction and valence band measurements were performed at a take-off angle of 88 degrees,

and the chamber pressure was approximately 10−9 torr. During measurement, the sample was

held at a constant bias of -15 V relative to the spectrometer in order to improve the low energy

photoelectron signal. Note that the samples have been exposed to atmosphere for a brief period

of time (~5 minutes) in order to apply carbon-tape conductive contacts to the films (to avoid

charging effects) prior to loading into the photoemission chamber.

Acknowledgments

This publication is based in part on work supported by Award KUS-11-009-21, made by the

King Abdullah University of Science and Technology (KAUST), by the Ontario Research Fund

Research Excellence Program, and by the Natural Sciences and Engineering Research Council.

Notes and references

1 J. Burschka, N. Pellet, S. J. Moon, R. Humphry-Baker, P. Gao, M. K. Nazeeruddin and M.

Grätzel, Nature, 2013, 499, 316–9.

2 M. Liu, M. B. Johnston and H. J. Snaith, Nature, 2013, 501, 395–398.

3 W. S. Yang, J. H. Noh, N. J. Jeon, Y. C. Kim, S. Ryu, J. Seo and S. I. Seok, Science, 2015,

aaa9272.

4 Z. K. Tan, R. S. Moghaddam, M. L. Lai, P. Docampo, R. Higler, F. Deschler, M. Price, A.

Sadhanala, L. M. Pazos, D. Credgington, F. Hanusch, T. Bein, H. J. Snaith and R. H. Friend,

Nat Nanotechnol, 2014, 9, 687–92.

5 Y.-H. Kim, H. Cho, J. H. Heo, T.-S. Kim, N. Myoung, C.-L. Lee, S. H. Im and T.-W. Lee,

Adv. Mater., 2014.

6 G. Xing, N. Mathews, S. S. Lim, N. Yantara, X. Liu, D. Sabba, M. Grätzel, S. Mhaisalkar

and T. C. Sum, Nat. Mater., 2014, 13, 476–480.

7 F. Deschler, M. Price, S. Pathak, L. E. Klintberg, D.-D. Jarausch, R. Higler, S. Hüttner, T.

Leijtens, S. D. Stranks, H. J. Snaith, M. Atatüre, R. T. Phillips and R. H. Friend, J. Phys.

Chem. Lett., 2014, 5, 1421–1426.

8 B. R. Sutherland, S. Hoogland, M. M. Adachi, C. T. O. Wong and E. H. Sargent, ACS Nano,

2014, 8, 10947–10952.

9 L. Dou, Y. (Micheal) Yang, J. You, Z. Hong, W.-H. Chang, G. Li and Y. Yang, Nat.

Commun., 2014, 5, 5404.

10 J. Luo, J.-H. Im, M. T. Mayer, M. Schreier, M. K. Nazeeruddin, N.-G. Park, S. D. Tilley, H.

J. Fan and M. Grätzel, Science, 2014, 345, 1593–1596.

11 S. D. Stranks, G. E. Eperon, G. Grancini, C. Menelaou, M. J. P. Alcocer, T. Leijtens, L. M.

Herz, A. Petrozza and H. J. Snaith, Science, 2013, 342, 341–344.

12 K. Yamada, K. Nakada, Y. Takeuchi, K. Nawa and Y. Yamane, Bull. Chem. Soc. Jpn., 2011,

84, 926–932.

13 E. Mosconi, A. Amat, M. K. Nazeeruddin, M. Grätzel and F. De Angelis, J. Phys. Chem. C,

2013, 117, 13902–13913.

14 S. Colella, E. Mosconi, P. Fedeli, A. Listorti, F. Gazza, F. Orlandi, P. Ferro, T. Besagni, A.

Rizzo, G. Calestani, G. Gigli, F. De Angelis and R. Mosca, Chem. Mater., 2013, 25, 4613–

4618.

15 E. Edri, S. Kirmayer, M. Kulbak, G. Hodes and D. Cahen, J. Phys. Chem. Lett., 2014, 5,

429–433.

16 B. Suarez, V. Gonzalez-Pedro, T. S. Ripolles, R. S. Sanchez, L. Otero and I. Mora-Sero, J.

Phys. Chem. Lett., 2014, 5, 1628–1635.

17 S. Colella, E. Mosconi, G. Pellegrino, A. Alberti, V. L. P. Guerra, S. Masi, A. Listorti, A.

Rizzo, G. G. Condorelli, F. De Angelis and G. Gigli, J. Phys. Chem. Lett., 2014, 5, 3532–

3538.

18 N. Kitazawa, Y. Watanabe and Y. Nakamura, J. Mater. Sci., 2002, 37, 3585 – 3587.

19 M. Zhang, H. Yu, M. Lyu, Q. Wang, J.-H. Yun and L. Wang, Chem. Commun., 2014, 50,

11727–11730.

20 J. H. Noh, S. H. Im, J. H. Heo, T. N. Mandal and S. I. Seok, Nano Lett., 2013, 13, 1764–

1769.

21 C. Bellabarba, J. Gonzalez and C. Rincon, Phys Rev B, 1996, 53, 7792–7796.

22 M. Saba, M. Cadelano, D. Marongiu, F. Chen, V. Sarritzu, N. Sestu, C. Figus, M. Aresti, R.

Piras, A. G. Lehmann, C. Cannas, A. Musinu, F. Quochi, A. Mura and G. Bongiovanni, Nat.

Commun., 2014, 5, 5049.

23 R. J. Elliott, Phys Rev, 1957, 108, 1384–1389.

24 V. D’Innocenzo, G. Grancini, M. J. P. Alcocer, A. R. S. Kandada, S. D. Stranks, M. M. Lee,

G. Lanzani, H. J. Snaith and A. Petrozza, Nat. Commun., 2014, 5.

25 P. Schulz, E. Edri, S. Kirmayer, G. Hodes, D. Cahen and A. Kahn, Energy Environ. Sci.,

2014, 7, 1377–1381.